Search index trimming

1. A method comprising:

prior to receiving, from a user device, a search query of one or more keywords searching for a relevant set of publications in a publication corpus, trimming, with one or more processors, candidate publications from a plurality of candidate publications to generate a trimmed plurality of candidate publications, the trimming based on (i) a machine-learned model of keywords to relevant publications; or (ii) historic user behavior comprising purchases, selections, and/or other interactions;

receiving the search query; and

in response to the search query, accessing, with the one or more processors, the trimmed plurality of candidate publications in the publication corpus.

2. The method of claim 1 , wherein the trimming is based on a machine learned model.

3. The method of claim 1 , wherein the trimming is based on historic user behavior data.

4. The method of claim 1 , further comprising:

for at least one keyword of a plurality of keywords including publication corpus keywords and potential search keywords, aggregating the plurality of candidate publications in the publication corpus.

5. The method of claim 1 , wherein the search query includes any of user profile data, session context data, and non-textual input that includes at least any of audio input, video input, or image input.

6. The method of claim 1 , further comprising:

responsive to receiving the search query, processing the trimmed plurality of candidate publications to search for the publication in the publication corpus.

7. The method of claim 6 , further comprising:

responsive to the processing the trimmed plurality of candidate publications, causing display of the relevant set of publications at the user device; and

after the causing display, receiving a selection signal originating from the user device, the selection signal indicating a selection from the at least one of the one or more closest matches.

8. A system comprising:

one or more processors and executable instructions accessible on a computer-readable medium that, when executed by the one or more processors, configure the one or more processors to at least perform operations comprising:

prior to receiving, from a user device, a search query of one or more keywords searching for a relevant set of publications in a publication corpus, trimming candidate publications from a plurality of candidate publications to generate a trimmed plurality of candidate publications, the trimming based on (i) a machine-learned model of keywords to relevant publications; or (ii) historic user behavior comprising purchases, selections, and/or other interactions;

receiving the search query; and

in response to the search query, accessing the trimmed plurality of candidate publications in the publication corpus.

9. The system of claim 8 , wherein the trimming is based on a machine learned model.

10. The system of claim 8 , wherein the trimming is based on historic user behavior data.

11. The system of claim 8 , wherein the operations further comprise:

for at least one keyword of a plurality of keywords including publication corpus keywords and potential search keywords, aggregating the plurality of candidate publications in the publication corpus.

12. The system of claim 8 , wherein the search query includes any of user profile data, session context data, and non-textual input that includes at least any of audio input, video input, or image input.

13. The system of claim 8 , wherein the operations further comprise:

responsive to receiving the search query, processing the trimmed plurality of candidate publications to search for the publication in the publication corpus.

14. The system of claim 13 , wherein the operations further comprise:

responsive to the processing the trimmed plurality of candidate publications, causing display of the relevant set of publications at the user device; and

after the causing display, receiving a selection signal originating from the user device, the selection signal indicating a selection from the at least one of the one or more closest matches.

15. A non-transitory machine-readable medium storing instructions that, when executed by one or more processors of a machine, cause the machine to at least perform operations comprising:

prior to receiving, from a user device, a search query of one or more keywords searching for a relevant set of publications in a publication corpus, trimming candidate publications from a plurality of candidate publications to generate a trimmed plurality of candidate publications, the trimming based on (i) a machine-learned model of keywords to relevant publications; or (ii) historic user behavior comprising purchases, selections, and/or other interactions;

receiving the search query; and

in response to the search query, accessing the trimmed plurality of candidate publications in the publication corpus.

16. The non-transitory machine-readable medium of claim 15 , wherein the trimming is based on a machine learned model.

17. The non-transitory machine-readable medium of claim 15 , wherein the trimming is based on historic user behavior data.

18. The non-transitory machine-readable medium of claim 15 , wherein the operations further comprise:

for at least one keyword of a plurality of keywords including publication corpus keywords and potential search keywords, aggregating the plurality of candidate publications in the publication corpus.

19. The non-transitory machine-readable medium of claim 15 , wherein the search query includes any of user profile data, session context data, and non-textual input that includes at least any of audio input, video input, or image input.

20. The non-transitory machine-readable medium of claim 15 , wherein the operations further comprise:

responsive to receiving the search query, processing the trimmed plurality of candidate publications to search for the publication in the publication corpus.